51 results about "Computational vision" patented technology

The present invention relates to the fields of multimedia information (picture and video), machine vision, and computational vision, and in particular to a method for fast video review, in which parameters are automatically assigned by the system to complete shot boundary extraction or similarity filtering, and then the extracted boundary frames are Combining a short video, and finally reviewing personnel or an automatic system to review the short video is a fast video review method, including step 1: input review type, uniformly input the review type of the video to be reviewed into the system preprocessing system; step 2 : Boundary lens extraction and similar frame filtering, step 3: video synthesis, using the open source tool ffmpeg (for example: ffmpeg‑fimage2‑i%d.jpg‑vcodec libx264‑rN test.mp4) command to generate a frame rate of N frames / s Step 4: Video review output, the generated video output with a frame rate of N frames / s is convenient for manual or system review.

The invention belongs to the technical field of intelligent robots, and discloses a vision acquisition and calibration method based on a multi-degree-of-freedom mechanical arm. The method comprises the following steps that a designated point on the mechanical arm is kept fixed, and the mechanical arm is respectively rotated for one circle around at least two coordinate axes of a coordinate systemtaking the designated point as an original point; coordinates of each corner point on a calibration plate in the rotating process are obtained; a circle center coordinate of the motion track of the calibration plate and a space offset vector from each corner point to the designated point under a vision acquisition coordinate system on a designated frame are calculated; the posture of the mechanical arm is kept unchanged from the designated point to the tail end of the mechanical arm, and the mechanical arm is respectively translated towards different planes; a plurality of frames of images aresampled for the calibration plate in the translation process, and meanwhile the coordinates of the tail end of the mechanical arm are obtained through the mechanical arm system; and the transformation relation between the vision acquisition coordinate system and the mechanical arm coordinate system is calculated. By adopting the method, the coordinate system transformation relation between visionacquisition equipment and the mechanical arm system can be calculated without accurately fixing the calibration plate without measurement equipment.